What is the UDEF?
The UDEF is a method of categorizing data. It assigns to every item of data an alphanumeric tag plus a simple textual description.

For example,"Purchase Order Number" in a government invoice is a commonly-encountered data element. It has a UDEF tag d.t.2_13.35.8 and description "Purchase . Order . DOCUMENT_Government . Assigned . IDENTIFIER"

How does the UDEF categorize data?
The UDEF uses the basic information systems model that is described in ISO 11179, in which every item of data is characterized as being an instance of a property of some object class, represented in some way. The UDEF provides hierarchical categorizations of object classes and properties. The UDEF tag for an item of data is obtained by putting together the tags for the object class and the property of which it is an instance.

Note that the categorization depends on the context in which the data is used. For example, information about a person in a corporate directory might have object class "Employee Person", but when that information is copied to a purchase order it might have object class "Approver Person". It is not possible to categorize data in the abstract, but each item of data has a unique and readily-determinable categorization in the context of a particular application.

Why is the UDEF needed?
It is needed to reduce the cost and improve the agility of enterprise information management.

A large enterprise typically has many different information stores and applications, which organize and categorize data in different ways. For information to flow between these stores and applications, either within an enterprise or between different enterprises, it must be analyzed for meaning and then transformed between different formats. This requires an understanding of the ways in which the data is organized and categorized. It is an expensive process, carried out by teams of programmers, who may have documented descriptions of existing software to help them, but generally have poor documentation of the information that they are transforming. The UDEF enables enterprises to categorize their information in a standard and consistent way. This will greatly reduce the cost of programming data transformations by providing the programmers with better documentation of the information that they are transforming. And it enables the development of automatic transformation methods that will reduce the cost further and will mean that new transformations can be introduced more quickly, in response to the business needs.

If changes occur to the UDEF trees, does this invalidate all my previously defined UDEF identifiers?
No. Changes will be made for new extensions, and to correct errors, but the meanings of existing UDEF definitions will never be changed. Therefore, a given UDEF identifier will not be invalidated by extensions to the UDEF.

The following principles apply when errors are detected in the UDEF trees.

• There shall be no changes to the tags. UDEF identifiers shall never be re-purposed.
• Informative descriptions may be added for qualifier terms.
• Editorial changes may be made to qualifiers and qualifier terms to correct spelling or syntax.
• Qualifier terms may be replaced by other terms that do not change the meaning but make the intention clearer.

UDEF tags may fall out of use as the concepts that they index become obsolete, but they will not be deleted or re-used for other concepts.

Who should be interested in the UDEF?
The UDEF is important to:

• People involved in data standards, such as STEP, X12, and XML variants;
• Enterprise architects;
• Systems analysts who are involved in building interfaces between systems; and
• Product designers working for vendors of major applications and meta data management products.

How is the UDEF different from an Ontology?
It is simpler. An information ontology will typically include an information categorization, and in addition will describe relationships between the classes of information that it identifies. Other than the hierarchical relationships between the object classes and properties in the UDEF trees, the UDEF does not do this.

There is no single, universal, information ontology. An enterprise may use many different ontologies, to meet the needs of its different information systems and applications. The World-Wide Web consortium has defined a language (OWL) for describing ontologies used in connection with the Web. (It assumes a similar object class/property information model to that of ISO 11179). Thus, while each ontology implies a method of categorizing data, ontologies do not provide a single universal categorization. The UDEF does provide such a categorization. By doing so, it can enable data categorized by different ontologies to be related.

How does the UDEF differ from data interchange standards such as STEP/PLCS and EDI?
There are many data interchange standards that are the basis of subject-area-specific information transfer syntaxes. A few examples are OAGIS, RosettaNet, STEP/PLCS, EIA-836, X12, EDIFACT, XBRL, and SWIFT. Unfortunately, there is substantial overlap and conflict between the various standards. For instance, each of the examples listed includes the concept of "address" but they do not use the same semantics. The UDEF is an indexing mechanism that is designed to enable information engineers to relate these standards to each other and to the intrinsic vocabularies used by the applications within each enterprise.

If there was no structure within words, using a dictionary would be impossible. Since words are made of letter symbols which follow a convention of ordering and other semantic considerations, you can take any word and look it up in either a dictionary or thesaurus. UDEF indexing offers essentially the same functionality.

As UDEF is merely a classification and indexing system for data elements, it does not conflict with information models such as STEP/PLCS or any other data standard. In this context, UDEF does not need to maintain the relationships between data elements from the original models, since it performs a rather different function.

The availability of mappings between the UDEF and standards such as STEP/PLCS and OAGIS will provide assistance to users, and will help to cut costs. Such mappings could be made available through the UDEF Registry or through repositories associated with it.

Why is The Open Group concerned with the UDEF?
The UDEF is a mechanism that can reduce enterprise costs associated with enabling information flow, and increase the agility of information management. Establishing the UDEF will be a significant contribution to realizing The Open Group's vision of Boundaryless Information Flow™.

How easy is it to categorize data using the UDEF?
A small amount of training (perhaps half a day for a competent information engineer) is needed to enable the user to determine which object class and property from the standard UDEF trees apply to any particular data item.

Is the UDEF complete?
The UDEF object and property trees currently include many of the common object classes and properties used by enterprises in the context of e-business. It is the aim of the UDEF Forum that theUDEF trees will be extended to include as nearly as possible all object classes and properties that are used by enterprises.

What are the primary objectives of the UDEF Project?
The main purpose of the Project is to establish the Universal Data Element Framework (UDEF) as the universally-used categorization system for data. The Project focuses on developing and maintaining the UDEF as an open standard, advocating and promoting it, putting in place a technical infrastructure to support it, implementing a Registry for it, and setting up education programs to train information professionals in its use.